Range-Doppler target image (RDI) signatures are often used in target recognition applications. In order to accurately extract target features from an RDI signature it is necessary for the range profiles of the RDI signature to have good time sidelobe performance (typically better than 20-25 dB below the level of the peak return for close in local peaks in the profile). If good time sidelobe performance is not achieved, “ghost” reflection sources can appear in the target signature, which can be interpreted as being part of the real target and severely degrade the feature extraction process. In order for the required time sidelobe performance to be achieved, the radar used to generate the RDI signature must be operating in its linear region. If the radar receiver receives unexpectedly high power radar return signals, for example from a target having a large radar cross-section or at a short range, it will become saturated and the time sidelobe performance and the consequent feature extraction performance will be degraded. There are various known methods for calibrating the time sidelobe performance of a radar system to enable the radar to be set to operate in its linear region, including using a remote point scatterer, such as a corner reflector, a remote delay line transponder, or a closed loop transponder to generate a calibration time sidelobe characteristic.